Is Precipitation Responsible for the Most Hydrological Model Uncertainty?

Rainfall-runoff modeling is highly uncertain for a number of different reasons. Hydrological processes are quite complex, and their simplifications in the models lead to inaccuracies. Model parameters themselves uncertain—physical because observations conceptual due limited identifiability. Furthermore, main model input—precipitation available high spatio-temporal variability. The quantification output uncertainty essential use. Most approaches used rainfall-runoff assign parameters. In this contribution, role precipitation investigated. Instead standard sensitivity analysis with respect input variations, it investigated what extent realistic fields could improve performance. Realistic defined as gridded realizations which reproduce observed values at observation locations, distribution spatial variability same observations. above conditions apply each time step. Through an inverse approach based on Random Mixing fulfilling reproducing discharge better than using traditional interpolated can be obtained. These show how much rainfall runoff may profit from remains parameter concept uncertainty. methodology applied two hydrological contrasting basis, SHETRAN HBV, three mesoscale sub-catchments Neckar basin Germany. Results that up 50% error attributed Further, inverting performance even neighboring catchments not considered explicitly.